Literature search log template for AI-assisted reviews

A literature search log records how papers were found. In an AI-assisted review, the log is even more important because search may involve databases, citation maps, natural-language AI queries, seed papers, and manual discovery.

Without a log, it becomes hard to explain why a paper entered the review, which queries were tested, which sources were searched, and how AI affected the source set.

This guide provides a search log template for AI-assisted literature reviews and explains what to record.

A literature search log is a record of search activities used to find sources for a review. It documents the route from research question to candidate papers.

A search log usually records:

• Search source.
• Date.
• Query.
• Filters.
• Number of results.
• Records saved.
• Notes.
• Follow-up actions.

For AI-assisted workflows, the log should also record AI search prompts, citation-chasing routes, and tool-assisted discovery steps.

A search log matters because AI search can be difficult to reconstruct later. Natural-language prompts, generated suggestions, and tool results may change over time.

The log helps you:

• Explain the search process.
• Avoid repeating failed searches.
• Track where papers came from.
• Compare query versions.
• Record AI-assisted discovery.
• Support PRISMA-style reporting.
• Audit whether key papers were missed.

The log turns search from a memory exercise into a method record.

For reporting, see PRISMA flow diagram with AI-assisted screening.

Every entry should answer what was searched, when it was searched, and what happened.

Include:

• Review question or subquestion.
• Search source or tool.
• Search date.
• Query or prompt.
• Filters or limits.
• Result count.
• Records saved.
• Source route.
• Reviewer.
• Notes.
• Next action.

If a query is changed, create a new entry rather than overwriting the old one. Failed searches are useful because they show what you tried.

Database searches should be recorded in enough detail that the search can be understood and, where possible, repeated.

Record:

• Database name.
• Platform if relevant.
• Full search string.
• Fields searched.
• Date searched.
• Date range.
• Language limits.
• Document type filters.
• Result count.
• Exported count.
• Notes about errors or changes.

If the same concept uses different terminology across databases, record the database-specific version.

For query building, see AI literature review search strings.

AI search queries should be logged as prompts or task descriptions. Natural-language searches can shape what papers are found, so they need a record.

Record:

• Tool used.
• Prompt or query.
• Date.
• Source type requested.
• Any limits given.
• Results inspected.
• Papers saved.
• How results were verified.
• Whether the query was revised.

For example, record whether you asked for peer-reviewed papers, recent studies, methods papers, or sources around a seed paper.

This avoids treating AI search as an invisible shortcut.

Citation chasing should be recorded as a source route from a seed paper. This applies to both backward and forward chasing.

Record:

• Seed paper.
• Direction: backward or forward.
• Tool or database.
• Date.
• Criteria for saving candidates.
• Papers saved.
• Notes on clusters or irrelevant branches.

Citation chasing often finds papers that keyword search misses. The log explains how those papers entered the review.

For the method, see backward and forward citation chasing.

Grey literature searches need extra source-type detail because the sources may not have the same metadata as journal articles.

Record:

• Organization or website searched.
• Search query or browsing route.
• Date.
• Source type.
• URL.
• Version or publication date.
• Inclusion reason.
• Quality note.
• Access note if needed.

Grey literature can be useful, but it needs careful labeling.

For criteria, see AI search for grey literature.

Use a field list like this:

Entry ID: A unique number or code.

Review question: Main question or subquestion.

Source type: Database, AI search, citation chasing, grey literature, manual search, or alert.

Tool or source: Name of the database, search engine, AI tool, or website.

Date searched: Date of the search.

Query or prompt: Exact search string or natural-language query.

Limits: Date, language, document type, field, source type, or other filter.

Results found: Number of records returned if available.

Records saved: Number of records exported or added to the candidate set.

Verification: How sources were checked.

Notes: Problems, revisions, missing terms, or next steps.

Use the log to understand where each record came from. This helps with deduplication, source-route reporting, and search improvement.

During screening, the log can show:

• Which query produced many irrelevant records.
• Which source found key papers.
• Which seed papers were productive.
• Which AI searches need verification.
• Which missing terms should be added.
• Which routes should be repeated later.

The log should remain connected to the screening record. Search and screening are separate stages, but they depend on each other.

Turn the advice into a repeatable workflow by defining the decision you need to make, the evidence required for that decision, and the record that will prove how the decision was made. In literature discovery, the problem is rarely one missing tool. The problem is usually that search, reading, checking, and writing happen in separate places without a shared rule.

Use a short operating routine:

• Name the review question or subquestion.
• Define the source set you are working from.
• Decide what counts as enough evidence for the next step.
• Apply the same criteria to every paper in that step.
• Mark uncertain cases instead of forcing a clean answer.
• Keep source locations for claims that may enter the final review.
• Review the workflow after each major search, screening, or writing session.

This routine keeps the work moving without making the review careless. It also gives supervisors, collaborators, and future you a way to understand why the source set changed.

Record the pieces that would be hard to reconstruct later. You do not need a diary of every click, but you do need enough detail to explain the path from question to source to claim.

For this topic, the most useful record usually includes queries, source routes, seed papers, result counts, and follow-up searches. Add the date, tool or source used, reviewer status, and next action. If AI assisted the step, write down what it helped with and what a human checked.

The record should distinguish discovery from evidence. A tool may help find a paper, but the paper itself must support the claim. A summary may help triage a source, but the original source should support any statement that appears in the literature review.

Before writing, check whether the workflow has produced usable evidence or only useful notes. Notes help you think. Evidence supports a sentence.

Ask:

• Which claim will this source support?
• Is the claim narrower than the evidence?
• Have methods, sample, outcome, or concept details been checked?
• Are limitations visible?
• Are conflicting papers handled rather than ignored?
• Is the citation real, current, and relevant?
• Can another reader understand how this source entered the review?

If the answer is unclear, keep the point in notes rather than moving it into the draft. This is the small pause that prevents AI-assisted research from becoming polished but weak writing.

WisPaper can support the search and source-set stages that a search log records. Deep Search, Scholar Agent, and Inspiration Discovery help researchers run natural-language academic searches, while paper cards show source labels, summaries, authors, publication details, and preview images.

Researchers can save or upload papers into My Library, helping keep candidate papers available for review after search. Library QA can answer questions based on the user's own library, which can help inspect saved papers before screening and extraction.

The search log itself should still record the query, date, source route, and human verification status. WisPaper helps with paper discovery and organization; the researcher keeps the method record.

Try WisPaper